The general optimal market area model with uncertain and nonstationary demand

Facility location problems involve determining the number of facilities to open, the location of facilities, and the market area that each facility will serve. While the environment is seldom stable, algorithms for solving models that incorporate change and uncertainty are rare and limited in scope. Consequently, the common approach in practice is to use algorithms for static models with perhaps some scenario analysis to accommodate change and uncertainty. We introduce an analytically tractable model that includes dynamic and uncertain demand. By studying how the model can be solved by solving a related model with stationary demand, we are able to derive guidelines on how to apply static analysis tools to problems with dynamic demand. While the analysis is contingent upon the underlying model, evidence on the model's robustness suggests that the results represent plausible guidelines for problems that extend well beyond the model.